Google is reportedly testing rewriting news story headlines using AI within its search results.
The move, which was first reported by The Verge, recreates headlines originally created by journalists using AI for a select number of articles in Google’s traditional search listings.
The search giant said the editing of story headlines aids how the story aligns with search intent, and that it wouldn’t only be titles from publications that would receive an AI-edit.
However, in some situations the overall meaning and tone of headlines has been altered, which raises concerns around spreading inaccurate news and misinformation.
Edited news headlines could also impact click-through rates for publications, who are already being impacted by readers navigating more to AI tools and less to website browsers to get the same information.
Google said the test was limited and hadn’t been fully rolled out yet, or determined if it would be.
Hallucinated Stories
This isn’t the first time altered AI-generated news headlines have come under scrutiny. In January 2025, Apple had to suspend a new AI feature that summarised news headlines because it often changed details of stories.
In one case, the feature falsely summarised a BBC News article, telling readers that the charged shooter that killed UnitedHealthcare CEO Brian Thompson, Luigi Mangione, had shot himself.
Articles from Sky News, the New York Times, the Washington Post were all amongst those incorrectly summarised.
AI tools often hallucinate – where the AI generates false, fabricated or nonsensical information – so when used for a purpose where a high degree of accuracy is needed, incorrect information has worse consequences, including the spread of fake news.
Readers often briefly scan the headline of a news story to gauge what has happened. If these are often inaccurate because of AI-edits, this could spell trouble for publications, and has the potential to damage publisher reputation and traffic.
